How prachi thinks
Decides and owns the call — moves before being told.
↳the log that shows itarticulate·curious·decisive
◈ Open to being sharpened
How prachi think, mapped · lit by their own logs · 7 showing
How you reason · the four moves
How you work · the qualities founders read for
Showing — it’s in their logs; tap for the receiptno %, no rank — evidence, not a score.
Milestones3 earned
Day 30 · Jul 31Day 20 · Jul 20Day 10 · Jul 09
The climbYou climbed.
The shape of your 32 entries
You climbed.
30 Jun → 6 Aug · valleys = stuck · peaks = shipped
Why this is the brag
Then you were stuck on icons!.
Now: stuck on running the expo app often leads….
Now: stuck on running the expo app often leads….
↑ You're stuck on a harder thing now, not the same one. That's real growth.
Every valley named in your own words. Nobody can fake a climb.
prachi · the character an interview can’t show
32 logs over 6 weeks — read from their own words.
+ Only points up3 of 4 traits shown · reasons across 4 of 4 moves
Character · what an interview can’t reach
✓Shown
Honest when it's not working
3 stuck and 0 hard days, logged plainly instead of hidden.
Proof · from their own logs
9 Jul
15 Jul
25 Jul
✓Shown
Adjusts — updates when the evidence changes
Changed course 1 time in the open, with the reason written in — not dug in.
Proof · from their own logs
2 Jul
◇Not yet
Lifts other people's work
Reads plenty (41 reactions) but hasn't yet weighed in on anyone's thinking — not a mark against them, just not yet on the page.
✓Shown
Makes the tools better
3 reports they filed got fixed — they improve what they use, not just what they're paid to.
Proof · reports they got fixed
4 Aug
4 Aug
4 Aug
The judgment behind it ↓
And the technical judgment holds up
Evidence8
Alternative10
Assumption4
Tradeoff7
About 2% of engineers on Proof reason across all four. Keep the tradeoffs coming. This is the easy read — an interview catches it too. The character above is the part it can’t.
The arc · they didn’t start here
Early
138
character whys
→
Lately
0
steady
Showed up · 23 days
3 stuck + 0 hard days, logged anyway. Nobody farms honest bad days — that’s why this counts and a streak doesn’t.
A hard day, in their words
running the expo app often leads to crashing...which sometimes gets resolved by tunneling (but often tunneling fails with ngrok error) someone who's worked on expo maybe could hel
Read prachi’s logs →sign in to connect with them
Recent logs30
Learned
Learning Log
The Feynman Technique - is a way of learning by explaining a concept in simple words. It helps identify gaps in understanding and improves retention.
Internalising - means understanding something so well that it becomes a natural part of your thinking and actions through practice.
A hypothesis - is a testable prediction or explanation based on observations. It is used to investigate ideas through evidence.
An assumption - is something accepted as true without proof. It is important to question assumptions to avoid incorrect conclusions.
Free write
Today's meeting log:
Discussed the importance of intellectual honesty and psychological safety within the team.
Emphasized documenting learnings, blockers, and meeting notes consistently.
AnimeshAug 4
Press further
Discussing a scenario would have made this sharper
Reciprocity
Sharpening means pressing on how someone thinks — and being open to the same yourself.
↓
One tap, once you’re in
Sign in — then you can sharpen any public log, and yours open in the same tap.
Sign in to sharpen →Free write
Last week's reflection:
I mostly held back. Every time I had to put an opinion out there, I'd freeze up. Proof's been a place to slowly practice getting out of that, talking daily, writing daily, having to actually show up instead of hiding. I'm not fully out of my shell yet, I still don't love how my voice sounds, but it's been a real, steady shift over just one week.
I used to treat "I can't communicate" as a fixed fact about myself. Practicing daily, with actual support from RK and this program, showed me it's trainable, I noticed real, small improvement just from repetition. That's not a belief anymore, it's something I've now watched happen to myself.
What I'd tell someone who's also struggling-
It won't make the fear disappear overnight. But it gives you an actual training ground, most people never get one for something like communication or confidence, they just avoid the discomfort indefinitely. Progress here is quiet and doesn't always feel like it's happening in the moment, but a good mentor helps you actually catch it happening, and that's worth more than it sounds like on paper.
Free write
Meeting notes with Surya (iraiga founder)
On curiosity & articulation:
Curiosity - explore, ask questions, seek new information and try to figure out things
Articulation matters just as much: clearly breaking a problem down in your own head before asking AI for help.
Question every AI output, LLMs hallucinate, don't take answers at face value.
On fundamentals:
Early-level focus should be fundamentals, time complexity, space complexity, not skipping ahead.
On software/engineering judgment:
Good software isn't about knowing the most, it's about making the best decisions with what you have.
On networking:
Community events and volunteering are real ways to build a network from scratch, not just cold outreach.
Free write
Today's meeting log:
Two traits matter more than raw technical skill right now: curiosity and high agency.
Curiosity -> leads you to ask the right questions.
High agency -> gives you the will to actually go ask them (and act on the answers), rather than waiting to be told.
Technical skills still matter, but they're not the differentiator anymore. Some technical aspects still matter a lot, like system design, distributed systems, networking, DSA, etc.
What differentiates now is code + judgment
p.s: acknowledging someone else's point ("I agree with X") counts as a form of appreciation, worth doing more consciously.
p.p.s: look into the 5 Whys technique, a way of repeatedly asking "why" on a problem to trace back to its actual root cause instead of stopping at the first symptom.
Free write
#10: Find a real problem first
Talked to 3 students, at different times, about problems they're facing. Without prompting each other, all three raised the same core issue: they're confused about internships and upskilling. Specifically, they don't know which skills actually matter, and don't know where or how to find good opportunities.
A second pattern also came up: a belief that better opportunities are out of reach because they're from a "lower-tier" college, an assumption made before even trying, not based on having actually been rejected.
So the real problem isn't a single thing, it's two, stacked: (1) a lack of clear direction on what to learn or where to look, and (2) a self-imposed belief that certain opportunities aren't available to them at all.
AnimeshJul 28
Learned something
Built
#9 Automate a task
Built an n8n automation that fetches news titles + links from defence.in 's RSS feed and sends them to me via Telegram.
It was a fun project to do!
Free write
#8 Make a page, show it to 5 people
Showed my RN developer portfolio to 5 people & asked for feedback. I'm still working on the projects section, but overall it was a positive experience. Got some suggestions and kind words back. Will be adding more projects and UI screenshots/previews.
Open to suggestions/reviews
Built
#7 build a screen log
built a mobile app screen of alarm clock on RN expo
Got stuck on flex layout, again. Need to learn Flexbox properly instead of trial-and-error-ing every time.
Assumed
#6 collecting data log:
Practiced data structuring using example data (roughly based on my experience attending winter school at IISc, BLR): food item prices across 5 canteens.
Visualized it as stacked bar chart to compare total price at a glance.
random lore: one specific canteen there was suspiciously cheap, like half-price-plus-a-rupee on literally everything compared to normal campus prices
Built
Built a small mobile app with expo to log how I'm spending my time.
It has easy to use ui- select time slot/s, pick a colour and give it a title.
next step would be to add persistence and one more screen to show weekly trends.
I've named the project "Where'd All The Time Go" or WATT-Go-- basically for people who easily lose track of time :)
Mistake
#5: Change a tutorial, log
Tried to build a currency converter mobile app today by myself first and got stuck on props.
Wrote:
```
const renderItem = (item: Currency) => { ... }
```
and passed renderItem to FlatList assuming FlatList would pass just the Currency item directly
FlatList internally calls renderItem with an object shaped like {item, index, separators}
so, item should be destructured out of it.
the correct approach would be
const renderItem = ({ item }: { item: Currency }) => { ... }
Free write
Today's meeting notes:
- we will be having recurring meets to discuss progress & blockers, if any
- decisions made in the meeting should be noted down and shared with team members
Importance of logging: not logging decisions, reasoning, changes can show up later during an escalation or failure, when someone has to go back and reconstruct "why did we do it this way" or "who decided this," and there's no trail.
It's very important and non-negotiable.
Learned
#3: How does WhatsApp work? (my mental model)
On the user level: UI (functional, code/logic) + local storage + server + network.
Local storage: actual message history, media, statuses, contacts cache.
Server: holds messages temporarily, just long enough to deliver them if the receiver is offline, then deletes them once delivered.
Network: WebSocket like persistent connection. WhatsApp being a native app just opens a raw TCP connection directly to the server.
Flow: user1 → server → user2. If user2 is offline, server holds the message until they reconnect.
Ticks:
- 1 tick = sent to server
- 2 ticks = delivered to user2's device
- blue ticks = user2 opened it, this is actually a second separate message traveling back (user2 → server → user1)
/n
Encryption: server only sees metadata (sender, receiver, timestamp) to route the message, never the payload. Payload is encrypted on-device before sending, only the receiving phone can decrypt it. Server never holds the key.
Free write
wrote a blog on rendering & re-rendering in React (useState, useEffect, DOM/virtual DOM/fiber tree).
Trying to explain it forced the understanding further.
Learned
a refresher about react native and rendering from this beginner friendly article on medium
why and how react native is "native" (it uses native ui components like android.view.View for android for example)
and what is it not: a WebView or a DOM-based framework like traditional web apps
Free write
I would fix expo go mobile app to have options in app to be compatible with different expo versions so i would not need to download different sdk's and uninstall and install again and again
Decided
I picked Ship, optimise if users complain over Optimise it now.
proof is a great example! users show up daily and even suggest changes. the app would be optimised regardless. starting should be as less friction as possible · I'll know I was right in 2 weeks if: i already have a product...if i get a real user- it would be even better!
AnimeshJul 21
Clear thinking
Stuck
running the expo app often leads to crashing...which sometimes gets resolved by tunneling (but often tunneling fails with ngrok error)
someone who's worked on expo maybe could help with the most stable version of expo or some other work around would be really helpful!
clearing the cache, using tunneling, tried different versions (i currently use the v55 expo go)
Learned
a brief intro about react native rendering architecture
Old Bridge & New architecture
Built
I built a small background changer app today which uses useState and a random hex generator function in React Native CLI
Decided
I picked Mostly sharpening fundamentals but also applying whenever anything interesting/relevant comes up. — not Sharpen fundamentals / interview prep or Apply to more roles.
strong foundation is necessary but some form of communication is also important. Interviews can be seen as kind of test to check how well do you actually understand the concepts. (sort of like rubber duck debugging :) · I'll know I was right in 2 weeks if: If i can explain what i've learnt & read and understand codebases (open-source for ex and start contributing in it)
Built
I built a react native (expo) mobile app to rate github repositories (ongoing)
Decided
I picked to revise the basics over hoarding concepts because revision help concepts stick longer
revision help concepts stick longer
Built
I built adopt a pet App with React Native (cli)
Decided
I picked native CLI over expo because abstraction without understanding creates knowledge gaps
abstraction without understanding creates knowledge gaps
Built
I built a react native (cli) password generator app
